Type II diabetes is a growing health concern for Latinos who not only have a higher incidence of the disease but also suffer great morbidity. At the same time due to poverty, language, low literacy and lack of continuity of care, this population is largely excluded from current diabetes education programs. To assist with this problem we propose to evaluate 1) a community-based, peer-led Spanish Diabetes Self-Management program based on self-efficacy theory, and 2) the effects of long-term tailored educational reinforcement offered by means of automated telephone disease management messages. Should this research be successful it will provide an evidenced based public health diabetes education model for use with Latino populations throughout the United States. The aims of the study are to: Implement and evaluate in a 6-month randomized trial an experimental community-based peer-led Spanish-language self-management program for patients with type 2 diabetes. Outcomes for evaluating the intervention's impact include health status (health-related quality-of-life and metabolic control) and health care utilization. Mediating outcomes are health behaviors and self-efficacy. 2. Determine the effectiveness of automated telephone disease management messages as an adjunction to the Spanish Diabetes Self-Management Program in a randomized, multiple assessment, and educational reinforcement trial. 3. Conduct a cost-benefit analysis of the above interventions. 4. Conduct an analysis to determine what baseline characteristics of participants best predict long-term, health-related quality of life and metabolic control as well as modeling the role of behaviors, self-efficacy and health distress on quality of life, metabolic and health care utilization outcomes.